Merge 4462d84461 into bfb2587697

pelican/contents.py

@@ -25,6 +25,25 @@ from pelican.urlwrappers import (Author, Category, Tag, URLWrapper)  # NOQA
 
 logger = logging.getLogger(__name__)
 
+try:
+    import html
+except ImportError:
+    # html.escape()/html.unescape() is since Python 3.2, do this for py2.7
+    # https://wiki.python.org/moin/EscapingHtml
+    from xml.sax.saxutils import escape, unescape
+
+    class html(object):
+        _html_escape_table = {'"': """,
+                              "'": "'"}
+        _html_unescape_table = {'"': '"',
+                                ''': "'"}
+
+        @classmethod
+        def escape(cls, v): return escape(v, cls._html_escape_table)
+
+        @classmethod
+        def unescape(cls, v): return unescape(v, cls._html_unescape_table)
+
 
 class Content:
     """Represents a content.
@@ -231,9 +250,9 @@ class Content:
 
     def _link_replacer(self, siteurl, m):
         what = m.group('what')
-        value = urlparse(m.group('value'))
+        value = urlparse(html.unescape(m.group('value')))
         path = value.path
-        origin = m.group('path')
+        origin = html.unescape(m.group('path'))
 
         # urllib.parse.urljoin() produces `a.html` for urljoin("..", "a.html")
         # so if RELATIVE_URLS are enabled, we fall back to os.path.join() to
@@ -333,7 +352,7 @@ class Content:
         # keep all other parts, such as query, fragment, etc.
         parts = list(value)
         parts[2] = origin
-        origin = urlunparse(parts)
+        origin = html.escape(urlunparse(parts))
 
         return ''.join((m.group('markup'), m.group('quote'), origin,
                         m.group('quote')))

pelican/tests/test_contents.py

@@ -362,28 +362,28 @@ class TestPage(TestBase):
         args['content'] = (
             'A simple test, with a '
             '<a href="|filename|article.rst'
-            '?utm_whatever=234&highlight=word">link</a>'
+            '?utm_whatever=234&amp;highlight=word">link</a>'
         )
         content = Page(**args).get_content('http://notmyidea.org')
         self.assertEqual(
             content,
             'A simple test, with a '
             '<a href="http://notmyidea.org/article.html'
-            '?utm_whatever=234&highlight=word">link</a>'
+            '?utm_whatever=234&amp;highlight=word">link</a>'
         )
 
         # combination
         args['content'] = (
             'A simple test, with a '
             '<a href="|filename|article.rst'
-            '?utm_whatever=234&highlight=word#section-2">link</a>'
+            '?utm_whatever=234&amp;highlight=word#section-2">link</a>'
         )
         content = Page(**args).get_content('http://notmyidea.org')
         self.assertEqual(
             content,
             'A simple test, with a '
             '<a href="http://notmyidea.org/article.html'
-            '?utm_whatever=234&highlight=word#section-2">link</a>'
+            '?utm_whatever=234&amp;highlight=word#section-2">link</a>'
         )
 
         # also test for summary in metadata
@@ -407,6 +407,21 @@ class TestPage(TestBase):
         self.assertEqual(p.summary, linked)
         self.assertEqual(p.custom, linked)
 
+        # SITEURL with characters that should be escaped
+        args['content'] = (
+            'A simple test, with a '
+            '<a href="|filename|article.rst'
+            '#highlight=&quot;word&quot;">link</a>'
+        )
+        content = Page(**args).get_content('http://notmyidea.org/'
+                                           '?app=blog&path=')
+        self.assertEqual(
+            content,
+            'A simple test, with a '
+            '<a href="http://notmyidea.org/?app=blog&amp;path='
+            '/article.html#highlight=&quot;word&quot;">link</a>'
+        )
+
     def test_intrasite_link_more(self):
         cls_name = '_DummyAsset'